《一週新論文》系列之2020年第9周：自然語言處理相關
本週重點關注：
- Microsoft: [2], [23], [40], [43], [76]
- Facebook: [36], [53], [78]
- Amazon: [21]
- Google: [10], [19], [34]
2020年2月28日
[1]. Generating Followup Questions for Interpretable Multi-hop Question Answering
鏈接 | https://arxiv.org/abs/2002.12344
作者 | Christopher Malon, Bing Bai
單位 | NEC Laboratories America
[2]. Few-shot Natural Language Generation for Task-Oriented Dialog
鏈接 | https://arxiv.org/abs/2002.12328
作者 | Baolin Peng, Chenguang Zhu, Chunyuan Li, Xiujun Li, Jinchao Li, Michael Zeng, Jianfeng Gao
單位 | Microsoft Research, Redmond
[3]. A Primer in BERTology: What we know about how BERT works
鏈接 | https://arxiv.org/abs/2002.12327
作者 | Anna Rogers, Olga Kovaleva, Anna Rumshisky
單位 | University of Massachusetts Lowell
[4]. Annotation of Emotion Carriers in Personal Narratives
鏈接 | https://arxiv.org/abs/2002.12196
作者 | Aniruddha Tammewar, Alessandra Cervone, Eva-Maria Messner, Giuseppe Riccardi
單位 | University of Trento; University of Ulm
Comments: To be published in LREC 2020
[5]. Improving cross-lingual model transfer by chunking
鏈接 | https://arxiv.org/abs/2002.12097
作者 | Ayan Das, Sudeshna Sarkar
單位 | IIT Kharagpur, India
[6]. Binarized PMI Matrix: Bridging Word Embeddings and Hyperbolic Spaces
鏈接 | https://arxiv.org/abs/2002.12005
作者 | Zhenisbek Assylbekov, Alibi Jangeldin
單位 | Nazarbayev University
[7]. Integrating Boundary Assembling into a DNN Framework for Named Entity Recognition in Chinese Social Media Text
鏈接 | https://arxiv.org/abs/2002.11910
作者 | Zhaoheng Gong, Ping Chen, Jiang Zhou
單位 | Harvard Business School; University of Massachusetts Boston; AI Strike
[8]. CrossWOZ: A Large-Scale Chinese Cross-Domain Task-Oriented Dialogue Dataset
鏈接 | https://arxiv.org/abs/2002.11893
作者 | Qi Zhu, Kaili Huang, Zheng Zhang, Xiaoyan Zhu, Minlie Huang
單位 | Tsinghua University
[9]. Analysis of diversity-accuracy tradeoff in image captioning
鏈接 | https://arxiv.org/abs/2002.11848
作者 | Ruotian Luo, Gregory Shakhnarovich
單位 | TTI-Chicago
[10]. Echo State Neural Machine Translation
鏈接 | https://arxiv.org/abs/2002.11847
作者 | Ankush Garg, Yuan Cao, Qi Ge
單位 | Google Research
[11]. Train Large, Then Compress: Rethinking Model Size for Efficient Training and Inference of Transformers
鏈接 | https://arxiv.org/abs/2002.11794
作者 | Zhuohan Li, Eric Wallace, Sheng Shen, Kevin Lin, Kurt Keutzer, Dan Klein, Joseph E. Gonzalez
單位 | UC Berkeley
[12]. Towards Zero-shot Learning for Automatic Phonemic Transcription
鏈接 | https://arxiv.org/abs/2002.11781
作者 | Xinjian Li, Siddharth Dalmia, David R. Mortensen, Juncheng Li, Alan W Black, Florian Metze
單位 | Carnegie Mellon University
Comments: AAAI 2020
[13]. Attacking Neural Text Detectors
鏈接 | https://arxiv.org/abs/2002.11768
作者 | Max Wolff
2020年2月27日
[14]. Marathi To English Neural Machine Translation With Near Perfect Corpus And Transformers
鏈接 | https://arxiv.org/abs/2002.11643
作者 | Swapnil Ashok Jadhav
Comments: 5 pages, 5 tables. This report is based on applied research work done at Dailyhunt
[15]. Using Distributional Thesaurus Embedding for Co-hyponymy Detection
鏈接 | https://arxiv.org/abs/2002.11506
作者 | Abhik Jana, Nikhil Reddy Varimalla, Pawan Goyal
Comments: Accepted in LREC 2020
[16]. Detecting Potential Topics In News Using BERT, CRF and Wikipedia
鏈接 | https://arxiv.org/abs/2002.11402
作者 | Swapnil Ashok Jadhav
Comments: 6 pages, 6 tables, 1 figure, 2 examples. This is a report based on applied research work conducted at Dailyhunt
[17]. End-to-End Entity Linking and Disambiguation leveraging Word and Knowledge Graph Embeddings
鏈接 | https://arxiv.org/abs/2002.11143
作者 | Rostislav Nedelchev, Debanjan Chaudhuri, Jens Lehmann, Asja Fischer
[18]. Object Relational Graph with Teacher-Recommended Learning for Video Captioning
鏈接 | https://arxiv.org/abs/2002.11566
作者 | Ziqi Zhang, Yaya Shi, Chunfeng Yuan, Bing Li, Peijin Wang, Weiming Hu, Zhengjun Zha
單位 | University of Science and Technology of China; University of Chinese Academy of Sciences
Comments: Accepted by CVPR 2020
[19]. Sparse Sinkhorn Attention
鏈接 | https://arxiv.org/abs/2002.11296
作者 | Yi Tay, Dara Bahri, Liu Yang, Donald Metzler, Da-Cheng Juan
單位 | Google AI
2020年2月26日
[20]. Semantic Relatedness for Keyword Disambiguation: Exploiting Different Embeddings
鏈接 | https://arxiv.org/abs/2002.11023
作者 | María G. Buey, Carlos Bobed, Jorge Gracia, Eduardo Mena
[21]. Language-Independent Tokenisation Rivals Language-Specific Tokenisation for Word Similarity Prediction
鏈接 | https://arxiv.org/abs/2002.11004
作者 | Danushka Bollegala, Ryuichi Kiryo, Kosuke Tsujino, Haruki Yukawa
單位 | University of Liverpool; Amazon
Comments: To appear in the 12th Language Resources and Evaluation (LREC 2020) Conference
[22]. A more abstractive summarization model
鏈接 | https://arxiv.org/abs/2002.10959
作者|Satyaki Chakraborty, Xinya Li, Sayak Chakraborty
[23]. MiniLM: Deep Self-Attention Distillation for Task-Agnostic Compression of Pre-Trained Transformers
鏈接 | https://arxiv.org/abs/2002.10957
作者 | Wenhui Wang, Furu Wei, Li Dong, Hangbo Bao, Nan Yang, Ming Zhou
單位 | Microsoft Research
[24]. Detecting Asks in SE attacks: Impact of Linguistic and Structural Knowledge
鏈接 | https://arxiv.org/abs/2002.10931
作者 | Bonnie J. Dorr, Archna Bhatia, Adam Dalton, Brodie Mather, Bryanna Hebenstreit, Sashank Santhanam, Zhuo Cheng, Samira Shaikh, Alan Zemel, Tomek Strzalkowski
單位 | Institute for Human and Machine Cognition; State University of New York; University of North Carolina, Charlotte; Rensselaer Polytechnic Institute
Comments: Accepted at AAAI 2020
[25]. KEML: A Knowledge-Enriched Meta-Learning Framework for Lexical Relation Classification
鏈接 | https://arxiv.org/abs/2002.10851
作者 | Chengyu Wang, Minghui Qiu, Jun Huang, Xiaofeng He
單位 | East China Normal University; Alibaba Group
[26]. Small-Footprint Open-Vocabulary Keyword Spotting with Quantized LSTM Networks
鏈接 | https://arxiv.org/abs/2002.10851
作者 | Théodore Bluche, Maël Primet, Thibault Gisselbrecht
單位 | Sonos Inc.
[27]. BERT Can See Out of the Box: On the Cross-modal Transferability of Text Representations
鏈接 | https://arxiv.org/abs/2002.10832
作者 | Thomas Scialom, Patrick Bordes, Paul-Alexis Dray, Jacopo Staiano, Patrick Gallinari
單位 | reciTAL; Sorbonne Universite; Criteo AI Lab
[28]. MuST-Cinema: a Speech-to-Subtitles corpus
鏈接 | https://arxiv.org/abs/2002.10829
作者 | Alina Karakanta, Matteo Negri, Marco Turchi
單位 | Fondazione Bruno Kessler; University of Trento
Comments: Accepted at LREC 2020
[29]. Label-guided Learning for Text Classification
鏈接 | https://arxiv.org/abs/2002.10772
作者 | Xien Liu, Song Wang, Xiao Zhang, Xinxin You, Ji Wu, Dejing Dou
[30]. Event Detection with Relation-Aware Graph Convolutional Neural Networks
鏈接 | https://arxiv.org/abs/2002.10757
作者 | Shiyao Cui, Bowen Yu, Tingwen Liu, Zhenyu Zhang, Xuebin Wang, Jinqiao Shi
單位 | Chinese Academy of Sciences; University of Chinese Academy of Sciences; Beijing University of Posts and Telecommunications
[31]. End-to-end Emotion-Cause Pair Extraction via Learning to Link
鏈接 | https://arxiv.org/abs/2002.10710
作者 | Haolin Song, Chen Zhang, Qiuchi Li, Dawei Song
單位 | Beijing Institute of Technology; University of Padua;
[32]. Multimodal Transformer with Pointer Network for the DSTC8 AVSD Challenge
鏈接 | https://arxiv.org/abs/2002.10695
作者 | Hung Le, Nancy F. Chen
單位 | Singapore Management University; Institute of Inforcomm Research (I2R)
Comments: Accepted at DSTC Workshop at AAAI 2020
[33]. Exploring BERT Parameter Efficiency on the Stanford Question Answering Dataset v2.0
鏈接 | https://arxiv.org/abs/2002.10670
作者 | Eric Hulburd
單位 | UC Berkeley
[34]. Differentiable Reasoning over a Virtual Knowledge Base
鏈接 | https://arxiv.org/abs/2002.10640
作者 | Bhuwan Dhingra, Manzil Zaheer, Vidhisha Balachandran, Graham Neubig, Ruslan Salakhutdinov, William W. Cohen
單位 | Carnegie Mellon University; Google Research
Comments: ICLR 2020
[35]. Parsing Early Modern English for Linguistic Search
鏈接 | https://arxiv.org/abs/2002.10546
作者 | Seth Kulick, Neville Ryant
單位 | University of Pennsylvania
[36]. On Feature Normalization and Data Augmentation
鏈接 | https://arxiv.org/abs/2002.11102
作者 | Boyi Li, Felix Wu, Ser-Nam Lim, Serge Belongie, Kilian Q. Weinberger
單位 | Cornell University; Cornell Tech; 3ASAPP Inc.; Facebook AI
[37]. Diversity-Based Generalization for Neural Unsupervised Text Classification under Domain Shift
鏈接 | https://arxiv.org/abs/2002.10937
作者 | Jitin Krishnan, Hemant Purohit, Huzefa Rangwala
單位 | George Mason University
[38]. Abstractive Snippet Generation
鏈接 | https://arxiv.org/abs/2002.10782
作者 | Wei-Fan Chen, Shahbaz Syed, Benno Stein, Matthias Hagen, Martin Potthast
單位 | Paderborn University; Leipzig University
Comments: Accepted by WWW 2020
[39]. Declarative Memory-based Structure for the Representation of Text Data
鏈接 | https://arxiv.org/abs/2002.10665
作者 | Sumant Pushp, Pragya Kashmira, Shyamanta M Hazarika
單位 | Central University of Jharkhand, India; National Institute of Technology, India;
[40]. Towards Learning a Generic Agent for Vision-and-Language Navigation via Pre-training
鏈接 | https://arxiv.org/abs/2002.10638
作者 | Weituo Hao, Chunyuan Li, Xiujun Li, Lawrence Carin, Jianfeng Gao
單位 | Duke University; Microsoft Research, Redmond
Comments: To appear at CVPR 2020.
2020年2月25日
[41]. Discriminative Adversarial Search for Abstractive Summarization
鏈接 | https://arxiv.org/abs/2002.10375
作者 | Thomas Scialom, Paul-Alexis Dray, Sylvain Lamprier, Benjamin Piwowarski, Jacopo Staiano
單位 | reciTAL; Sorbonne Universite; CNRS
[42]. Multilingual Twitter Corpus and Baselines for Evaluating Demographic Bias in Hate Speech Recognition
鏈接 | https://arxiv.org/abs/2002.10361
作者 | Xiaolei Huang, Linzi Xing, Franck Dernoncourt, Michael J. Paul
單位 | University of Colorado Boulder; University of British Columbia; Adobe Research
Comments: Accepted at LREC 2020
[43]. Low-Resource Knowledge-Grounded Dialogue Generation
鏈接 | https://arxiv.org/abs/2002.10348
作者 | Xueliang Zhao, Wei Wu, Chongyang Tao, Can Xu, Dongyan Zhao, Rui Yan
單位 | Peking University; Microsoft
Comments: Published in ICLR 2020
[44]. Improving BERT Fine-Tuning via Self-Ensemble and Self-Distillation
鏈接 | https://arxiv.org/abs/2002.10345
作者 | Yige Xu, Xipeng Qiu, Ligao Zhou, Xuanjing Huang
單位 | Fudan University; Huawei Technologies Co., Ltd.
[45]. Semi-Supervised Speech Recognition via Local Prior Matching
鏈接 | https://arxiv.org/abs/2002.10336
作者 | Wei-Ning Hsu, Ann Lee, Gabriel Synnaeve, Awni Hannun
單位 | MIT; Facebook AI Research
[46]. Word Embeddings Inherently Recover the Conceptual Organization of the Human Mind
鏈接 | https://arxiv.org/abs/2002.10284
作者 | Victor Swift
單位 | University of Toronto
[47]. Fixed Encoder Self-Attention Patterns in Transformer-Based Machine Translation
鏈接 | https://arxiv.org/abs/2002.10260
作者 | Alessandro Raganato, Yves Scherrer, Jörg Tiedemann
單位 | University of Helsinki
[48]. Learning to Select Bi-Aspect Information for Document-Scale Text Content Manipulation
鏈接 | https://arxiv.org/abs/2002.10210
作者 | Xiaocheng Feng, Yawei Sun, Bing Qin, Heng Gong, Yibo Sun, Wei Bi, Xiaojiang Liu, Ting Liu
單位 | Harbin Institute of Technology; Tencent AI Lab
Comments: accepted by AAAI2020
[49]. Predicting Subjective Features from Questions on QA Websites using BERT
鏈接 | https://arxiv.org/abs/2002.10107
作者 | Issa Annamoradnejad, Mohammadamin Fazli, Jafar Habibi
單位 | Sharif University of Technology
[50]. GRET: Global Representation Enhanced Transformer
鏈接 | https://arxiv.org/abs/2002.10101
作者 | Rongxiang Weng, Haoran Wei, Shujian Huang, Heng Yu, Lidong Bing, Weihua Luo, Jiajun Chen
單位 | Nanjing University; Alibaba Group
Comments: Accepted by AAAI 2020
[51]. Do Multi-Hop Question Answering Systems Know How to Answer the Single-Hop Sub-Questions?
鏈接 | https://arxiv.org/abs/2002.09919
作者 | Yixuan Tang, Hwee Tou Ng, Anthony K.H. Tung
單位 | National University of Singapore
[52]. Fill in the BLANC: Human-free quality estimation of document summaries
鏈接 | https://arxiv.org/abs/2002.09836
作者 | Oleg Vasilyev, Vedant Dharnidharka, John Bohannon
單位 | Primer Technologies Inc.
[53]. Unsupervised Question Decomposition for Question Answering
鏈接 | https://arxiv.org/abs/2002.09758
作者 | Ethan Perez, Patrick Lewis, Wen-tau Yih, Kyunghyun Cho, Douwe Kiela
單位 | Facebook AI Research; New York University; University College London; CIFAR
[54]. Exploiting Typed Syntactic Dependencies for Targeted Sentiment Classification Using Graph Attention Neural Network
鏈接 | https://arxiv.org/abs/2002.09685
作者 | Xuefeng Bai, Pengbo Liu, Yue Zhang
單位 | Zhejiang University; Westlake University; Harbin Institute of Technology
[55]. Incorporating Effective Global Information via Adaptive Gate Attention for Text Classification
鏈接 | https://arxiv.org/abs/2002.09673
作者 | Xianming Li, Zongxi Li, Yingbin Zhao, Haoran Xie, Qing Li
單位 | Ant Financial Services Group; City University of Hong Kong; Lingnan University; Hong Kong Polytechnic University
[56]. Machine Translation System Selection from Bandit Feedback
鏈接 | https://arxiv.org/abs/2002.09646
作者 | Jason Naradowsky, Xuan Zhang, Kevin Duh
單位 | Preferred Networks; Johns Hopkins University
[57]. Markov Chain Monte-Carlo Phylogenetic Inference Construction in Computational Historical Linguistics
鏈接 | https://arxiv.org/abs/2002.09637
作者 | Tianyi Ni
單位 | Arizona State University
[58]. Data Augmentation for Copy-Mechanism in Dialogue State Tracking
鏈接 | https://arxiv.org/abs/2002.09634
作者 | Xiaohui Song, Liangjun Zang, Yipeng Su, Xing Wu, Jizhong Han, Songlin Hu
單位 | Chinese Academy of Sciences; Baidu Inc.
[59]. Efficient Sentence Embedding via Semantic Subspace Analysis
鏈接 | https://arxiv.org/abs/2002.09620
作者 | Bin Wang, Fenxiao Chen, Yuncheng Wang, C.-C. Jay Kuo
單位 | University of Southern California
[60]. “Wait, I’m Still Talking!” Predicting the Dialogue Interaction Behavior Using Imagine-Then-Arbitrate Model
鏈接 | https://arxiv.org/abs/2002.09616
作者 | Zehao Lin, Xiaoming Kang, Guodun Li, Feng Ji, Haiqing Chen, Yin Zhang
單位 | Zhejiang University; Alibaba Group;
[61]. Emergent Communication with World Models
鏈接 | https://arxiv.org/abs/2002.09604
作者 | Alexander I. Cowen-Rivers, Jason Naradowsky
單位 | Huawei R&D London; Preferred Networks
Comments: NeurIPS Workshop on Emergent Communication
[62]. Training Question Answering Models From Synthetic Data
鏈接 | https://arxiv.org/abs/2002.09599
作者 | Raul Puri, Ryan Spring, Mostofa Patwary, Mohammad Shoeybi, Bryan Catanzaro
單位 | Nvidia; Rice University
[63]. Extracting and Validating Explanatory Word Archipelagoes using Dual Entropy
鏈接 | https://arxiv.org/abs/2002.09581
作者 | Yukio Ohsawa, Teruaki Hayashi
單位 | The University of Tokyo
[64]. Modelling Latent Skills for Multitask Language Generation
鏈接 | https://arxiv.org/abs/2002.09543
作者 | Kris Cao, Dani Yogatama
單位 | DeepMind
[65]. KBSET – Knowledge-Based Support for Scholarly Editing and Text Processing with Declarative LaTeX Markup and a Core Written in SWI-Prolog
鏈接 | https://arxiv.org/abs/2002.10329
作者 | Jana Kittelmann, Christoph Wernhard
單位 | Martin-Luther Universit¨at Halle-Wittenberg, Germany
Comments: To appear in DECLARE 2019 Revised Selected Papers
[66]. Uncertainty based Class Activation Maps for Visual Question Answering
鏈接 | https://arxiv.org/abs/2002.10309
作者 | Badri N. Patro, Mayank Lunayach, Vinay P. Namboodiri
單位 | Indian Institute of Technology
[67]. Rhythm, Chord and Melody Generation for Lead Sheets using Recurrent Neural Networks
鏈接 | https://arxiv.org/abs/2002.10266
作者 | Cedric De Boom, Stephanie Van Laere, Tim Verbelen, Bart Dhoedt
[68]. Leveraging Code Generation to Improve Code Retrieval and Summarization via Dual Learning
鏈接 | https://arxiv.org/abs/2002.10198
作者 | Wei Ye, Rui Xie, Jinglei Zhang, Tianxiang Hu, Xiaoyin Wang, Shikun Zhang
單位 | Peking University; University of Texas at San Antonio
Comments: Published at The Web Conference (WWW) 2020, full paper
[69]. FONDUE: A Framework for Node Disambiguation Using Network Embeddings
鏈接 | https://arxiv.org/abs/2002.10127
作者 | Ahmad Mel, Bo Kang, Jefrey Lijffijt, Tijl De Bie
單位 | Ghent University
[70]. Emosaic: Visualizing Affective Content of Text at Varying Granularity
鏈接 | https://arxiv.org/abs/2002.10096
作者 | Philipp Geuder, Marie Claire Leidinger, Martin von Lupin, Marian Dörk, Tobias Schröder
[71]. Deep Multimodal Image-Text Embeddings for Automatic Cross-Media Retrieval
鏈接 | https://arxiv.org/abs/2002.10016
作者 | Hadi Abdi Khojasteh, Ebrahim Ansari, Parvin Razzaghi, Akbar Karimi
[72]. Automata for Hyperlanguages
鏈接 | https://arxiv.org/abs/2002.09877
作者 | Borzoo Bonakdarpour, Sarai Sheinvald
單位 | Iowa State University
[73]. Sketching Transformed Matrices with Applications to Natural Language Processing
鏈接 | https://arxiv.org/abs/2002.09812
作者 | Yingyu Liang, Zhao Song, Mengdi Wang, Lin F. Yang, Xin Yang
單位 | University of Wisconsin-Madison; Princeton University; University of California, Los Angeles; University of Washington
Comments: AISTATS 2020
2020年2月24日
[74]. Is Aligning Embedding Spaces a Challenging Task? An Analysis of the Existing Methods
鏈接 | https://arxiv.org/abs/2002.09247
作者 | Russa Biswas, Mehwish Alam, Harald Sack
[75]. Refinement of Unsupervised Cross-Lingual Word Embeddings
鏈接 | https://arxiv.org/abs/2002.09213
作者 | Magdalena Biesialska, Marta R. Costa-jussà
[76]. Learning Dynamic Knowledge Graphs to Generalize on Text-Based Games
鏈接 | https://arxiv.org/abs/2002.09127
作者 | Ashutosh Adhikari, Xingdi Yuan, Marc-Alexandre Côté, Mikuláš Zelinka, Marc-Antoine Rondeau, Romain Laroche, Pascal Poupart, Jian Tang, Adam Trischler, William L. Hamilton
單位 | University of Waterloo; Microsoft Research; Charles University; Vector Institutel; MILA; McGill University
[77]. On the impressive performance of randomly weighted encoders in summarization tasks
鏈接 | https://arxiv.org/abs/2002.09084
作者 | Jonathan Pilault, Jaehong Park, Christopher Pal
單位 | Element AI; MILA; CIFAR
Comments: Accepted to ACL 2019 SRW. First two authors contributed equally
[78]. Accessing Higher-level Representations in Sequential Transformers with Feedback Memory
鏈接 | https://arxiv.org/abs/2002.09402
作者 | Angela Fan, Thibaut Lavril, Edouard Grave, Armand Joulin, Sainbayar Sukhbaatar
單位 | Facebook AI Research
[79]. Crowdsourced Collective Entity Resolution with Relational Match Propagation
鏈接 | https://arxiv.org/abs/2002.09361
作者 | Jiacheng Huang, Wei Hu, Zhifeng Bao, Yuzhong Qu
單位 | Nanjing University; RMIT University
Comments: Accepted by the 36th IEEE International Conference on Data Engineering (ICDE 2020)
[80]. Language as a Cognitive Tool to Imagine Goals in Curiosity-Driven Exploration
鏈接 | https://arxiv.org/abs/2002.09253
作者 | Cédric Colas, Tristan Karch, Nicolas Lair, Jean-Michel Dussoux, Clément Moulin-Frier, Peter Ford Dominey, Pierre-Yves Oudeyer
想要了解更多的自然語言處理最新進展、技術乾貨及學習教程，歡迎關注微信公衆號“DestinedAI”或掃描二維碼添加關注。
